OpenRouter 模型 ↗ · 2026-11-10
thinkingmachines/inkling-2-small
Model release OpenRouter 模型 Source language: Chinese
OpenRouter 于 2026-11-10 收录该模型(收录时间不等于模型正式发布日期);上下文:1,048,576 tokens;输入价:$0.3000/百万 tokens;输出价:$1.0000/百万 tokens;定位(来源描述):Inkling 2 Small is an open-weight multimodal mixture-of-experts reasoning model from Thinking…
Full text
The free Inkling 2 Small endpoint is only available for use with agentic harnesses. Do not upload any confidential information or personal data (e.g., voices and images of people's faces). Your usage of this free endpoint, including prompts and outputs, is logged and used to improve Thinking Machines Lab's models, products, and services. The logged session data will be disassociated from your account and other persistent identifiers before being used for these purposes.
By using this free endpoint, you agree to the TML Free Research API Terms of ServiceOpens in new tab. For more information about Thinking Machines Lab's data processing practices, see this Privacy NoticeOpens in new tab.
Thinking Machines: Inkling 2 Small
thinkingmachines/inkling-2-small
Model weights
Compare
Inkling 2 Small is an open-weight multimodal mixture-of-experts reasoning model from Thinking Machines Lab, with 17B active parameters out of 594B total. It is the first model in the second Inkling generation. It was trained with reinforcement learning on long-horizon coding tasks across many agent harnesses, and is suited for agentic coding, autonomous research, agentic knowledge work and retrieval-augmented generation, with English and general multilingual support. It accepts text and image input (no audio), supports tool calling and variable reasoning effort, and has a 1M-token context window.
Modalities
In / out price
$0.15 / $0.50per 1M
Context
1.0M
Released
Nov 9, 2026
Compare
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).
Latency / throughput
$0.300$1.00$0.03000.48s263 tok/s
98.78%
$0.300$1.00$0.03000.80s257 tok/s
99.67%
$0.300$1.00$0.03000.50s212 tok/s
99.59%
$0.150$0.50$0.03000.86s109 tok/s
99.60%
5% off
$0.30$0.285$1.00$0.95$0.03$0.02850.65s153 tok/s
99.56%
$0.300$1.00$0.06001.47s191 tok/s
96.62%
ProviderInput / Output /M
Input $0.30 / Output $1.00
263 tok/s98.78% uptime
Input $0.30 / Output $1.00
257 tok/s99.67% uptime
Input $0.30 / Output $1.00
212 tok/s99.59% uptime
Input $0.15 / Output $0.50
109 tok/s99.60% uptime
Input $0.285 / Output $0.95
5% off
153 tok/s99.56% uptime
Input $0.30 / Output $1.00
191 tok/s96.62% uptime
Pricing
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Weighted average
Input
$0.04387/M tokens
Output
$0.9491/M tokens
Inkling 2 Small — Price history
Price history
No effective price data for this selection
$0.0355$0.50095.4%13.6%
$0.0406$1.00096.1%12.3%
$0.0418$1.00095.6%47.0%
$0.0455$1.00094.2%17.4%
$0.0613$0.95087.2%6.7%
$0.0790$1.00092.1%3.0%
ProviderInput / Output /M
Token share
Effective in $0.0355 / Effective out $0.500
Token share
13.6%
Effective in $0.0406 / Effective out $1.000
Token share
12.3%
Effective in $0.0418 / Effective out $1.000
Token share
47.0%
Effective in $0.0455 / Effective out $1.000
Token share
17.4%
Effective in $0.0613 / Effective out $0.950
Token share
6.7%
Effective in $0.0790 / Effective out $1.000
Token share
3.0%
Performance
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Throughput
263tok/s
P50, best across providers
Latency
0.48s
P50, best provider
Reasoning effortLatency / throughput
AutoExacto benchmarks
GPQA DiamondTAU-BenchDeepInfra90.2%70.7%Nebius Token Factory91.4%65.3%Together91.4%64.0%Prime Intellect91.1%63.3%Modal90.9%63.3%
+1 more providers
Uptime
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Uptime (3d)The model was reachable. Request routed to a provider.
99.99%
Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.
99.74%
Availability over the last 3 days
Last 72 hours
Availability 99.74%
3 Days Ago2 Days AgoYesterdayNow
Availability over the last 24 hours
OpenRouter Availability
99.92%
Without Routing
81.55%
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.
Benchmarks
Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.
Benchmark score summary for Thinking Machines: Inkling 2 Small (Artificial Analysis and Vals AI)
SourceBenchmarkScore
Artificial AnalysisInkling 2 Small Intelligence Index30.6
Artificial AnalysisInkling 2 Small HLE44.1%
Artificial AnalysisInkling 2 Small AA-LCR83.3%
Artificial AnalysisInkling 2 Small GDPval-AA56.5%
Artificial AnalysisInkling 2 Small SciCode52.3%
Artificial AnalysisInkling 2 Small AA-Omniscience Accuracy33.8%
Artificial AnalysisInkling 2 Small AA-Omniscience Non-Hallucination Rate73.3%
Vals AIInkling 2 Small BioMysteryBench70.0%
Vals AIInkling 2 Small CyberBench v1.172.7%
Vals AIInkling 2 Small EMB v1.169.5%
Vals AIInkling 2 Small Finance Agent v2.167.5%
Vals AIInkling 2 Small Vals Finance Index61.5%
Vals AIInkling 2 Small HLAB-Vals-1.121.7%
Vals AIInkling 2 Small Vals Legal Index73.9%
Vals AIInkling 2 Small Legal Research Bench46.6%
Vals AIInkling 2 Small MysteryMechanism25.2%
Vals AIInkling 2 Small ProgramBench0.5%
Vals AIInkling 2 Small SAGE88.6%
Vals AIInkling 2 Small Vals Science Index32.7%
Vals AIInkling 2 Small SRE Bench1.9%
Vals AIInkling 2 Small SWE-bench Multimodal43.1%
Vals AIInkling 2 Small Tax Agent Bench38.9%
Vals AIInkling 2 Small Terminal-Bench 4.012.6%
Vals AIInkling 2 Small Terminal-Bench-Science2.9%
Vals AIInkling 2 Small Vals Index56.1%
Vals AIInkling 2 Small Vibe Code Bench v1.181.5%
Apps
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
1.
Claude Code
Claude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.
12.1Btokens
2.
omp
new
2.75Btokens
3.
OpenHands
AI coding agent that can run commands, browse the web, call APIs
2.24Btokens
4.
pi
There are many coding agents, but this one is yours.
1.59Btokens
5.
Hermes Agent
Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.
1.13Btokens
Rest of OpenRouter
33.5B
Top 5 Apps
5.31B
Activity
Token volume and request traffic to this model over time.
API
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
Not currently available through EU in-region routing. In-region routing docs
Frequently asked questions
Inkling 2 Small is an open-weight multimodal mixture-of-experts reasoning model from Thinking Machines Lab, with 17B active parameters out of 594B total. It is the first model in the second Inkling generation.
Inkling 2 Small costs $0.15/M input tokens and $0.50/M output tokens, with separate rates for Cache Read at $0.03/M tokens.
Inkling 2 Small has a 1,048,576 token context window. It supports up to 1,048,576 completion tokens.
Yes. Inkling 2 Small accepts tools for function calling. It also supports structured outputs via a JSON schema in response_format.
Inkling 2 Small accepts text and images as input and returns text.
Inkling 2 Small is served by 6 providers on OpenRouter: Prime Intellect, DeepInfra, Modal, Nebius Token Factory, Together, and Cloudflare. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.
Inkling 2 Small was released on November 9, 2026.